Weather windows

[temptress]

So reading a message from a friend of mine I was considering weather forecasting and weather windows. My friend just recently aborted a trip across Biscay and headed to the Channel Islands. This ultimately resulted in assistance by the RNLI (another story for him to tell)

We are currently at anchor in Tonga looking for a window between depressions to get the 1000 miles or so to Vanuatu. Can't see anything in the forecast that allows this so we are going to go to Fiji instead some 300 miles away in the right direction.

I am still getting here thinking that no matter where you are sailing you are always a hostage to the forecast. When we were I Europe we had good and reliable access to weather information yet we still seemed to always been worried about the options. In the Pacific information is much harder to come by but still as contradictory.

With all this in mind how do you decide what is a suitable forecast and what is not?

 

[KellysEye]

Offshore you have to take any weather that comes along but if you have a marine SSB there are Pacific forecasts which you can get as voice, fax or mailsail with a Pactor modem. We also had a Sat C terminal which gives forecasts anywhere in the world. We woudn't set off in a gale but if it but on passage it was fine, in a gale over Biscay we got gusts up to 50 knots.

 

[Ric]

Get a subscription to Predictwind. It has a departure planning option which will compare your options up to four days ahead and then give a readout of likely sea-states, wind, percentage time spent reaching, upwind, downwind etc plus a lot of other information. The subscription fee is well worth it - it is tiny compared to the cost of sailing your yacht to the other side of the world and then sitting in an unpleasant and avoidable weather system smashing your boat.

 

[macd]

I don't think temptress is asking for forecast-type recommendations. He's asking what sort of decision process follows the forecast: "With all this in mind how do you decide what is a suitable forecast and what is not?"

 

[RobbieW]

Paraphrasing Frank, consistency is king. If using GRIB, look at a 7-10 day output over several days, preferably with 24hr separation between model outputs. If those models are giving a consistent picture over time theres a fair chance the actual weather will be something similar!

 

[dolphin]

Marine Weather Bulletin
Issued from the National Weather Forecasting Centre Nadi
at 7:30pm on Friday the 15th of July 2016

Situation:
A southeast wind flow prevails over Fiji.

Forecast to midnight tomorrow for Fiji waters:

Southeast winds 15 to 20 knots.
Mpderate to rough seas.
Moderate southerly swells.

Further outlook: Southeast winds 15 to 20 knots. Moderate to rough
seas. Moderate southerly swells.

more or less this is the weather for next week

sent PM

 

[KellysEye]

>I don't think temptress is asking for forecast-type recommendations. He's asking what sort of decision process follows the forecast: "With all this in mind how do you decide what is a suitable forecast and what is not?"

OK, we had a heavy displacement steel ketch so the best wind to leave in was 20 to25 knots which we needed to get up to hull speed.

 

[macd]

Paraphrasing Frank, consistency is king. If using GRIB, look at a 7-10 day output over several days, preferably with 24hr separation between model outputs. If those models are giving a consistent picture over time theres a fair chance the actual weather will be something similar!

+1
Forecasts that have fluctuated markedly over a period should be treated with the utmost caution. I'd suggest that's by far the biggest single factor in trusting them.

 

[temptress]

I don't think temptress is asking for forecast-type recommendations. He's asking what sort of decision process follows the forecast: "With all this in mind how do you decide what is a suitable forecast and what is not?"

+1
Forecasts that have fluctuated markedly over a period should be treated with the utmost caution. I'd suggest that's by far the biggest single factor in trusting them.

The issue is not availability of forecasts but the conflicting forecasts. Different sites do not agree. That sail I prefer using the models from SQUID.

What I was hoping for here was a discussion on different decision making processesn to get some outside input. We get stuck in our ways sometimes.

 

[macd]

I'd add to the above by suggesting that the degree of difference between different models/types of forecast is equally suggestive of reliability. (Note 'diferent models/types': not different websites, since many will use the same data and models and thus generate the same forecasts: only the presentation is likely to differ.) Of the grib-type forecasts, Passageweather is probably the most convenient for this, since it readily dispays forecasts for a variety of models. On the other hand the scale is often quite small.

If the impending weather is dominated by low-pressure systems, then the degree of concurrence will usually be smaller: that's the nature of the beast (and often the price of sailing outside the traditional seasons.)

There are probably worse fates than having to 'waste' time in Tonga. Or Fiji. We'll form an orderly queue

 

[franksingleton]

The issue is not availability of forecasts but the conflicting forecasts. Different sites do not agree. That sail I prefer using the models from SQUID.

What I was hoping for here was a discussion on different decision making processesn to get some outside input. We get stuck in our ways sometimes.

Use one model. The leaders are ECMWF, NOAA GFS, UK, JMA. The only one readily and freely available in GRIB form is the GFS. Most commercial organisations either simply regurgitate the GFS or, in some cases, eg Predictwind,I believe, run another model themselves. I would never bother with these, nor would I ever pay for them. If they run models then their starting point – their initial data analysis can be no better than the GFS and is probably not so good.

Weather models are developed by or on behalf of National Weather Services. There is much interchange of knowledge and experiences so that no model is likely to be better than any other. Differences in models on particular occasions are more likely to be due to slight differences in the starting analysis. As far as I know, no organisation other than the major national or international centres has the ability, knowledge and experience to process the basic data to provide the initial analysis. Such organisations as Theyr.com, Predictwind, WindGuru, Windfinder etc can only start with the GFS T=0 output. This is only avaiable to them on a 0.25 degree grid. That is degraded from the grid used by the GFS.

More importantly, detailed forecasts by commercial concerns do not (again as far as I know) use any detailed data. Claims for precise forecasts are untenable, partly for reasons of lack of analysis detail. The other factor is the short lifetimes of small scale weather features. You cannot predict something of size around20-40 km in any useful manner. Lifetimes are a few hours only. By the time observational data had been collected, processed to start a forecast, the forecast run and results broadcast such size features would have disappeared.

Apply the consistency check method with the GFS. Overall you will get better results than using any commercial provider. On specific occasions, one of the latter may come up trumps but the GFS will do better in general.

In Reeds Weather Handbook, I take a pragmatic realistic approach. That is the consequence of a long lifetime in meteorology and a little more as a sailor.

 

[Harry Brown]

What are the platform options to obtain GFS?

 

[franksingleton]

What are the platform options to obtain GFS?

With a laptop PC or Apple Mac you can use zyGrib for direct download. See http://www.zygrib.org/ and/or http://weather.mailasail.com/Franks-Weather/Zygrib-Ftp-Grib-Service

If you can only access the Internet by email, then Saildocs has an email service. See http://weather.mailasail.com/Franks-Weather/Saildocs-Free-Grib-Files.


With tablets or smartphones there are several Apps. See http://weather.mailasail.com/Franks-Weather/Grib-Files-Apps. Some of these apps limit the number of days ahead. Weather4D (IOS or Android) is one that lets you access forecasts 10 days ahead. It can also get data via Saildocs email. That lets you look at the forecasts on several viewers depending on your preference. One such viewer is the iNavX chart plotter.

All these services give GFS output and all are free, courtesy of the US government. Tablet apps are available at low cost. You can try a few and see which presentation you prefer.